A Convolution Neural Network (CNN) From Scratch

This was written using basic python to deeply understand the forward and backward propagation of a neural network.

Usage

Install dependencies:

$ pip install -r requirements.txt

Then, run it with no arguments:

$ python cnn.py
$ python cnn_torch.py

Name		Name	Last commit message	Last commit date
Latest commit History 34 Commits
weights		weights
CNN_Python.ipynb		CNN_Python.ipynb
CNN_Python_CIFAR.ipynb		CNN_Python_CIFAR.ipynb
CNN_Python_MNIST_and_CIFAR.ipynb		CNN_Python_MNIST_and_CIFAR.ipynb
CNN_Torch.ipynb		CNN_Torch.ipynb
CNN_Torch_MNIST_and_CIFAR.py		CNN_Torch_MNIST_and_CIFAR.py
LICENSE		LICENSE
README.md		README.md
cnn.py		cnn.py
cnn_keras.py		cnn_keras.py
conv.py		conv.py
conv_filters.npy		conv_filters.npy
conv_filters_3d.npy		conv_filters_3d.npy
data.csv		data.csv
fc.py		fc.py
filters.npy		filters.npy
linear_biases.npy		linear_biases.npy
linear_filters.npy		linear_filters.npy
maxpool.py		maxpool.py
python_cnn.py		python_cnn.py
relu.py		relu.py
requirements.txt		requirements.txt
soft_fc.py		soft_fc.py
softmax.py		softmax.py
softmax_test.py		softmax_test.py
test.py		test.py
test2.py		test2.py
yolov2_tiny.py		yolov2_tiny.py